如何同时否定多个过滤条件答案

【问题标题】：How to negate multiple filter condition at same time如何同时否定多个过滤条件
【发布时间】：2022-01-24 15:20:57
【问题描述】：

示例：我有一个类似的数据框

我想用否定过滤多个条件firstname == "James" & lastname == "Smith" or firstname == "Robert" & lastname == "Williams"

我需要的输出应该是我正在使用类似的东西，但它不起作用

df = df.filter(~(df.firstname == "James") & (df.lastname == "Smith")|~(df.firstname == "Robert") & (df.lastname == "Williams"))

【问题讨论】：

请在问题正文中说明您的情况，至于您要过滤什么
我已经修改好了吗？我知道获得所需输出的逻辑，但我想使用数据框过滤器对多个条件使用否定

标签： python pandas dataframe apache-spark pyspark

【解决方案1】：

A|B 的否定是(~A)&(~B)。

所以试试这个：

df = df.filter((~(firstname == "James" & lastname == "Smith")) & (~(firstname == "Robert" & lastname == "Williams")))

因为C&D的否定是(~C)|(~D)，你可以进一步简化你的过滤条件为

df = df.filter((firstname != "James" | lastname != "Smith") & (firstname != "Robert" | lastname != "Williams"))

【讨论】：

【解决方案2】：

这里的 OR 条件没有给出正确的输出。我们必须将其更改为 AND

df_new = (df
          .filter(~((F.col("firstname") == "James") & (F.col("lastname") == "Smith"))
                  & ~((F.col("firstname") == "Robert") & (F.col("lastname") == "Williams"))
                 )
         )

结果如下 -

【讨论】：

括号中的否定表达式我试过了，我也在寻找否定过滤器内多个条件的答案
我已经更新了我的答案，检查

【解决方案3】：

您必须对整个条件应用否定。


data = [("James","","Smith","36636","M",3000),
    ("Michael","Rose","jim","40288","M",4000),
    ("Robert","","Williams","42114","M",4000),
    ("Maria","Anne","Jones","39192","F",4000),
    ("Jen","Mary","Brown","60563","F",-1)]

df = spark.createDataFrame(data, ("firstname", "middlename", "lastname", "id", "gender", "salary", ))

(df.filter(~(((df.firstname == "James") & (df.lastname == "Smith")) |
             ((df.firstname == "Robert") & (df.lastname == "Williams"))
            )
          )
   .show())

输出

+---------+----------+--------+-----+------+------+
|firstname|middlename|lastname|   id|gender|salary|
+---------+----------+--------+-----+------+------+
|  Michael|      Rose|     jim|40288|     M|  4000|
|    Maria|      Anne|   Jones|39192|     F|  4000|
|      Jen|      Mary|   Brown|60563|     F|    -1|
+---------+----------+--------+-----+------+------+

【讨论】：